Learning Geometry-Aware Nonprehensile Pushing and Pulling with Dexterous Hands
This addresses the problem of manipulating difficult-to-grasp objects for robotics, offering a novel method that leverages dexterous hands for more versatile nonprehensile actions.
The paper tackles nonprehensile manipulation with dexterous robotic hands by proposing Geometry-aware Dexterous Pushing and Pulling (GD2P), which generates hand poses via contact-guided sampling and trains a diffusion model to predict viable poses, achieving scalable motion generation applicable to different hand morphologies in real-world experiments.
Nonprehensile manipulation, such as pushing and pulling, enables robots to move, align, or reposition objects that may be difficult to grasp due to their geometry, size, or relationship to the robot or the environment. Much of the existing work in nonprehensile manipulation relies on parallel-jaw grippers or tools such as rods and spatulas. In contrast, multi-fingered dexterous hands offer richer contact modes and versatility for handling diverse objects to provide stable support over the objects, which compensates for the difficulty of modeling the dynamics of nonprehensile manipulation. Therefore, we propose Geometry-aware Dexterous Pushing and Pulling(GD2P) for nonprehensile manipulation with dexterous robotic hands. We study pushing and pulling by framing the problem as synthesizing and learning pre-contact dexterous hand poses that lead to effective manipulation. We generate diverse hand poses via contact-guided sampling, filter them using physics simulation, and train a diffusion model conditioned on object geometry to predict viable poses. At test time, we sample hand poses and use standard motion planners to select and execute pushing and pulling actions. We perform extensive real-world experiments with an Allegro Hand and a LEAP Hand, demonstrating that GD2P offers a scalable route for generating dexterous nonprehensile manipulation motions with its applicability to different hand morphologies. Our project website is available at: geodex2p.github.io.